A retrospective early test of Pinus sylvestris seedlings grown at wide and dense spacing

Stored seeds from 30 open-pollinated families that had been field tested for 28 years provided a retrospective opportunity to assess the efficiency of early testing in Scots pine (Pinus sylvestris L.). Seedlings height at the end of first (HGP1), second (HGP2) and third growth period (HGP3), and in the third growth period, leader length (LEADER3), basal stem diameter (DIA3), number of stem units (NSU3), mean stem unit length (MSUL3), aboveground fresh weight (AGFWT3), stem dry weight (STDWT3), and shoot and needle dry weight (SHNDWT3) were studied under wide and dense spacing. Families differed at the 1 or 0.1% level of probability. Differences between spacing were significant at the 0.01% level of probability for all traits measured in the third growth period. Families interacted significantly with spacing for HGP2, HGP3, LEADER3, AGFWT3, SHNDWT3 and MSUL3. Interactions were due to rank changes and the interaction component of variance varied from 26 to 96% of the family variance. Estimates of hi2 over wide and dense spacing ranged between 0.23 ± 0.11 (HGP2) and 0.51 ± 0.14 (NSU3). Spacing did not alter the variance structure, hi2 for each trait under wide or dense spacing were similar. However, spacing affected the phytotron-field genetic correlations. The average of 80 genetic correlation estimates was 0.52 under wide spacing and 0.29 under dense spacing. Averaged over eight field traits, the efficiency of indirect juvenile selection for the field traits indirect-direct selection percentage ranged from -15% for NSU3 under dense spacing to 99% for HGP3 under wide spacing.